4/1/2022 Aaron Seidlitz, Illinois CS 4 min read
The first-year Illinois CS professor’s research strengths in wireless sensing and networking helped identify the accuracy issues that hamper current localization efforts at extended range.
Written by Aaron Seidlitz, Illinois CS
Beginning in 2020, before joining Illinois CS, first-year professor Elahe Soltanaghai overcame several challenges to continue researching and developing Millimetro as a postdoctoral researcher at Carnegie Mellon University.
The product introduces what she describes as “an ultra-low-power tag” developed “in the context of autonomous driving to efficiently localize roadside infrastructure such as lane markers and road signs, even if obscured from view, where visual sensing fails.”
Her devotion to wireless sensing and networking with applications in cyber-physical systems, and the Internet of Things, started to take shape as a PhD student at the University of Virginia and expanded while at CMU.
Over the course of the last two years, she and her workgroup developed, first, the proof of concept for Millimetro, and have since expanded its viability as a product. She continues the work now at Illinois CS.
The paper that became the result of their work was also accepted at the 27th Annual International Conference on Mobile Computing and Networking (MobiCom ’21), which is the top-tier conference in wireless and mobile computing. In addition, the team received the first place in the student research competition. Here's the teaser video the group utilized at the conference for Millimetro.
The group will continue work developing Millimetro as Soltanaghai builds her research group here at Illinois CS.
“The motivation for Millimetro stems from a combination of two things,” Soltanaghai said. “We wanted to do something impactful; that is really important to me. While it is, of course, nice to receive recognition in the form of publication, I wanted this to accomplish something more – and we all believe Millimetro could be in the hands of millions of users one day.”
It’s clear from the response both in academic and industry settings, that Soltanaghai’s product could improve current localization efforts at extended ranges for applications with anything from autonomous vehicles, drones, mobility robotics to even the Metaverse – which she considers to be more realistic now in the hands of Facebook and with the need to create a “physical, virtual person” in the age of remote work.
Up to this point, Soltanaghai’s work with sensors provided insight into its weakness. As her abstract for Millimetro points out, “current ultra-low-power localization systems struggle to operate accurately at extended ranges under strict latency requirements.”
This means that the sensors used for projects like autonomous vehicles work less effectively at long distance and due to various interruptions.
Millimetro found a solution in the form of “re-using existing automotive radars that operate at mmWave frequency where plentiful bandwidth is available to ensure high accuracy and low latency.”
“What I’ve come to believe is the biggest impact of a system like Millimetro is the scalability,” Soltanaghai said. “Of course, we’ve envisioned this in use outside on the roads for autonomous driving, but also inside buildings or anywhere that is going to use radars.
“Really, we are at the edge of building a system inspired by a key question. How can we interact with everything that is not necessarily network connected?”
From the moment she and this workgroup – which included two PhD students, Akarsh Prabhakara and Artur Balanuta, and two faculty from CMU, Anthony Rowe and Swarun Kumar – came up with the concept in 2020, they worked through a pandemic that caused them to shift initial work from the lab setting at CMU to their homes.
It’s this kind of impact that drives the professor, even as her research group overcame complications from the realities of working during the COVID-19 pandemic.
Soltanaghai said that it took about nine months to reach the final product they wanted. Along the way, she conducted experiments on the balcony of her house.
Every time the sensitive hardware received electrostatic discharge by someone accidentally touching it, she would have to reset things. Others experienced similar frustrations, but they also were able to reconvene back on campus as one of the first groups back in lab spaces at CMU.
“We separated because of COVID right after the project started, but I was so fortunate to work with a group of people equally passionate about the idea. We all worked hard together,” Soltanaghai said.